HOW TO EVALUATE YOUR TRUANCY REDUCTION PROGRAMNational Center for School Engagement September 2005An initiative of the Colorado Foundation for Families and Children 303 E. 17th Avenue, Suite 400 Denver, CO 80203 303/837-8466 www.schoolengagement.
orgHow to Evaluate Your Truancy Reduction ProgramPlan Aheado It is best to plan for evaluation at the same time that you design your program. o Set aside money for evaluation in your initial budget application. o Do not give up on evaluation if you did not plan it at the beginning. Start now!Create a Logic Model of Your Programo Decide what outcome(s) you want. Some possibilities are improved attendance school-wide; improved attendance among participating students; improved grades among participating students; increased feelings of attachment to school among participating students, or their families, or the student body in general. o What program components will be most likely to achieve your goals o Be sure your program components and your designated outcomes match. In other words, if all your interventions are aimed at a small subset of students with serious attendance problems, you may not see measurable outcomes across the entire student body.
But that will not mean that your program is not working ??“ it means your evaluation is not well designed.Know What Kind of Evaluation You Needo Process evaluation, also known as implementation evaluation, looks at the way in which a program is set up and is operating. It is particularly relevant when an established program already shown to be effective is being copied in a new location.
Unless the program is implemented as intended, one cannot expect to achieve the same good results. Process evaluation lends itself to both qualitative and quantitative data collection. o Outcome evaluation measures the effects of a program once it has been established. A program will not likely be ready for an outcome evaluation until it has been in operation for some time ??“ generally a year or more.
If begun too early, an outcome evaluation will likely show no results and may unnecessarily dampen enthusiasm for a potentially good program. o Cost-benefit analysis compares the cost of a program as measured in dollars to the outcomes of that program, also measured in dollars. If benefits are greater than costs, then the program may be said to pay off. Sometimes it is difficult to put a price on benefits; cost-effectiveness analysis may be used to compare the cost of a program as measured in dollars, to benefits measured in something other than dollars.
For example, how many additional high school credits are earned as aresult of a $50,000 truancy reduction program Cost-effectiveness analysis is most useful when comparing multiple programs.Think About Your Data and Data Collection Methodso Data come in two general forms, quantitative and qualitative. Each provides a distinct purpose and you must consider what kind of data will be best used. o Studies of school attendance lend themselves easily to quantitative analysis, but the most thorough evaluations include both. o School records can provide data on outcome measures such as grades, class credits, disciplinary referrals, and attendance, which can be correlated with race, gender and age. o Students, parents and school personnel can be surveyed. Be careful how you formulate survey questions; borrowing from other surveys is the best way to get meaningful questions.
Provide clear directions so that respondents understand how to complete the survey. Always have a professional review your survey before administering it; remember junk in ??“ junk out. o Interviews can include structured questions in which respondents select from a set of designated responses; these data are quantitative.
Interviews may also include open-ended questions that allow respondents to answer any way they please; these questions provide qualitative data. Direct quotations should be taken. o Focus groups bring together a group of people ??“ usually between 5 and 10 ??“ to discuss their experiences. Questions are open-ended and discussion is encouraged.
Focus groups should be tape recorded and transcribed for analysis. Focus groups provide qualitative data.Know How to Collect and Analyze Quantitative Data??? ??? ??? Quantitative data are numeric; they can be counted and measured. Obvious examples include student grades on a 4-point scale, days or class periods attended or missed, and class credits earned. Quantitative data such as attendance, grades and credits may be collected from the school??™s administrative records. Quantitative data on the number of students served by a TRP, or the number of students who received tutoring or a family service referral, may be collected from the records of a TRP??™s social worker. Be sure to keep track of these things.
Create a system for recording social worker interventions on an on-going basis and keep up to date.???Most survey data are also quantitative. Administering a survey is the most practical way to collect data from a large sample of respondents. The following is an example of a quantitative survey question.How helpful was the tutoring program in motivating you to attend class regularly Please circle your answer. 1 2 3 4 5 Not at All Somewhat Very Helpful Helpful Helpful ??? Several computer programs are useful aids in analyzing quantitative data. SPSS and SAS are statistical software packages capable of sophisticated analysis. You will need someone competent with statistics and familiar with the programs to use them.
You may need to contract with a qualified consultant for your data analysis. Microsoft Excel is adequate for very simple record keeping and analyses, but not sufficient for advanced statistical analysis. The Truancy Reduction Application Interface (TRAIN) is a web-based data collection and analysis program available from the National Center for School Engagement for a technical assistance fee.
No statistical or programming experience is needed.Know How to Collect and Analyze Qualitative Datao Qualitative data are non-numeric. They are descriptive text passages, observations, field notes, records or documents, audio or video files.
o Qualitative data can come from interviews and focus groups, and sometimes from open-ended questions on written surveys. Collecting and analyzing qualitative data is time-consuming, and therefore expensive. It is most practical to collect qualitative data from a relatively small sample of respondents.
o While quantitative data are best suited for answering the ???what??? question, qualitative data are best for answering the ???how??? and ???why??? questions. Quantitative data may tell you a program did not work, but you will not know why it did not work without some qualitative data. o It is best to tape record and transcribe interviews and focus groups to record actual responses. o The following is an example of a qualitative interview question. What effect did the tutoring program have on your class attendance o Several computer programs are useful aids in analyzing large amounts of qualitative data. NVivo and Atlas-ti allow users to import text files, code electronically, and gather all selections with the same code for analysis. The programs themselves arerelatively user friendly, but training in qualitative data analysis is needed to know what to do with your data.What is Longitudinal Evaluation and When Would I use Ito Longitudinal evaluations measure changes that occur within the same group of students over time.
o They require taking a ???baseline??? measurement, in other words, measuring the outcome variable(s) before the intervention begins (or in its very early stages if need be), and comparing it to a second measurement of the same variable(s) taken at the end of the program. This is also called ???pre??? and ???post??? or a time series design. o Both quantitative and qualitative data may be collected longitudinally, though a longitudinal study is generally thought of as having a quantitative component. Example of quantitative baseline data: (number of class periods skipped in the month prior to TRP intervention) (total number of class periods in the month prior to TRP intervention). Compare the above truancy rate to the comparable figure for the last month of TRP participation: (number of class periods skipped in the last month of TRP participation) (total number of class periods during that month). Was there improvement How much Was the change large enough to be meaningful o Does improvement last beyond the end of the program Follow participants over time to find out.
What happens three months after the end of the program Six months One year o Note: if attendance typically varies according to the month of the year you should consider comparing, for example, a student??™s attendance rate in the November prior to TRP participation, with that student??™s attendance rate in the November following program participation.What are Control Groups and Experimental Groupso The best way to be sure that a program is having an effect is to compare students who participate in the TRP to students who do not. o Students in the experimental group receive a treatment that is under evaluation ??“ in this case, they participate in a TRP. Students in the control group, sometimes called the comparison group, do not participate.
Researchers are then able tocompare changes in attendance among students in the experimental group to changes in attendance among students in the control group over time. o Why is a control group important If you discover only marginal improvement in attendance from one year to the next among TRP participants, you may be discouraged. However, truancy typically worsens as children get older. If attendance among your control group declines significantly over the same time period, your program may be having a greater impact than it appears based solely on participant behavior.
o In order for the comparison to be valid, experimental and control groups must be as similar as possible in as many factors as possible. First and foremost, they should have similar patterns of absences; but other variables to consider include gender, racial/ethnic background, age, economic background, and school characteristics among others. o The best way to ensure that experimental and control groups are similar is to randomly assign students to each group. Ethics must be considered. Is it ethical to deny a truant child the services of a TRP If your budget is limited and you cannot serve all the eligible students, perhaps it is. If your budget is large enough to serve everyone and you have good cause to believe that the program will be effective because you have copied a proven program, perhaps it is not. Do not fail to consider ethical issues. o A good alternative source of a control group is to study students in a nearby school that does not offer a TRP.
Be sure the school serves a similar population of students. o If your evaluation budget cannot support a control group, do not give up. Survey, interview, or conduct focus groups among TRP participants to understand what effect the program had from their perspective. If they began attending class more regularly, why did they do so Did it have anything to do with the program Or was it mostly due to other factors If their attendance did not improve, whyThe National Center for School Engagement (NCSE) is an initiative of The Colorado Foundation for Families and Children (CFFC). NCSE strives to build a network of key stakeholders who share the belief that improving school attendance and school attachment promotes achievement and school success. NCSE was established as a result of more than a decade of educational research about youth out of the educational mainstream conducted by CFFC.
The impact of this work has been the development of significant investments of state funds to reduce suspensions expulsions and truancy. Over five years ago, CFFC began working with the OJJDP, US Department of Justice to assist in the planning and implementation of pilot demonstration projects across the country. As projects developed, CFFC became the national evaluator of this five-year truancy demonstration project. The culmination of ten years of program experience and research has identified truancy and school engagement as the centerpiece of NCSE??™s work to improve outcomes for youth who are at the greatest risk of school failure and delinquency. We are national leaders in applying research to help communities prevent and reduce truancy.
Co-Authors:Joanna Zorn Heilbrunn Heather McGillivaryNational Center for School Engagement c/o Colorado Foundation for Families and Children 303 E. 17th Avenue, Suite 400 Denver, CO 80203 (303) 837-8466 www.schoolengagement.org